cenggokhan opened a new issue #18030: Model load problem in 64 bit devices
URL: https://github.com/apache/incubator-mxnet/issues/18030
 
 
   I compiled amalgamation for armeabi-v7a and arm64-v8a, I am trying to test 
model 
https://github.com/flyingzhao/mxnet_VanillaCNN/blob/master/vanilla-0001.params 
and 
https://github.com/flyingzhao/mxnet_VanillaCNN/blob/master/vanilla-symbol.json
   devices with arch armeabi-v7a loads model without any problem. But arm64-v8a 
devices can not load model. Exception:
   org.dmlc.mxnet.MxnetException: [21:53:21] jni/../mxnet_predict-all.cc:18084: 
Check failed: from.shape() == to->shape() operands shape mismatch
   When I try same model with python this exception occurs as arm64 device. 
Exception:
   Shape inconsistent, Provided = [100,576], inferred shape=(100,256)
   My python code:
   
   import mxnet as mx
   from mxnet import model
   import numpy as np
   import cv2
   
   img=cv2.imread("test.jpg")
   img=cv2.resize(img,(40,40))
   m=model.FeedForward.load('vanilla',1,num_batch_size=1)
   b=np.zeros((1,3,40,40),dtype=np.float32)      #change chanels into 
shape[1,3,40,40]
   b[0,0,:,:]=img[:,:,2]
   b[0,1,:,:]=img[:,:,1]
   b[0,2,:,:]=img[:,:,0]
   
   points=m.predict(b)
   print(points)
   
   Please help.
   

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

Reply via email to